Vermelding onderdeel organisatie

February 1, 2012

1

Chapter 15: Failure modes and optimisation

ct5308 Breakwaters and Closure Dams

H.J. Verhagen

Faculty of Civil Engineering and Geosciences Section Hydraulic Engineering

Sri Lanka, Kudawella Tsunami damage of breakwater

2004

February 1, 2012 2

What is the most important element of a breakwater or closure dam ??

• the element which is the most expensive one

• the section which is the most costly one

• the element which is the most unreliable one

• the element which is the most sensitive to variations in the boundary conditions

February 1, 2012 3

failure modes for dike-type structures

Failure of breakwater by earthquake

February 1, 2012 4

failure modes for a rubble mound (Burcharth, 1992)

February 1, 2012 5

Failure modes for a monolithic breakwater

February 1, 2012 6

rock fill overflow dam failure modes

February 1, 2012 7

fault tree

February 1, 2012 8

fault tree for closure dam (cross section)

February 1, 2012 9

fault tree for closure dam (equipment)

February 1, 2012 10

fault tree for construction planning

February 1, 2012 11

equipment utilisation in relation to fault tree

February 1, 2012 12

The Dilemma

• A strong and heavy breakwater does not require maintenance … but is very expensive to construct

• A light breakwater is much cheaper to construct … but requires a lot of maintenance

February 1, 2012 13

Wave climate

Wave Height H

(m)

Probability of

Exceedance

(times per annum)

4 1.11

5 1.58*10-1

5.2 8.4*10-2

5.5 7.62*10-2

5.8 3.8*10-2

6 2.47*10-2

6.5 7.35*10-3

7.15 3.0*10-3

7.25 2.63*10-3

7.8 9.0*10-4

7.98 8.0*10-4

8.7 1.5*10-4

wave exceedance

0

2

4

6

8

10

0.00010.0010.010.11

exceedance (times per year)

wa

ve

he

igh

t (m

)

February 1, 2012 14

development of damage

Actual Wave Height H Damage in % of armour layer

H < Hnd 0

Hnd < H < 1.3Hnd 4

1.3Hnd < H < 1.45 Hnd 8

H > 1,45 Hnd Collapse

February 1, 2012 15

Cost of construction

Design wave height

Hnd

Initial cost

breakwater

“C”

Initial cost Amour

Layer

“A”

(m) (€) per running meter (€) per running meter

4 13900 5280

5 15220 6600

5.5 15900 7280

6 16540 7920

Initial cost for armour units: € 1320 * Hd

for core: € 8620

February 1, 2012 16

annual risk

1 < H < 1.3 Hnd

n = 4% damage

1.3 Hnd < H < 1.45 Hnd

n = 8% damage

H > 1.45 Hnd

Collapse

Hnd

p w p.w p w p.w p w p.w

(m) (1/year) (€) (€/year) (1/year) (€) (€/year) (1/year) (€) (€/year)

4 1.02 420 430 4.6 10-2

860 40 3.8 10-2

13900 530

5 1.5 10-1

530 80 4.7 10-3

1060 5 2.6 10-3

15220 40

5.5 7.4 10-2

580 40 2.2 10-3

1160 - 8 10-4

15900 10

6 2.4 10-2

630 15 7.5 10-4

1260 - 1.5 10-4

16540 3

p probability of occurrence of the wave height

w cost of repair of the armour layer (2nA)

respectively cost of replacement (C)

February 1, 2012 17

average annual risk

s = (p.w) Hnd

Full repair of partial

damage

Only repair of serious

damage(>8%)

No repair of partial

damage

(m) (€ per year) (€ per year) (€ per year)

4 1000 570 530

5 125 45 40

5.5 50 10 10

6 18 3 3

February 1, 2012 18

capitalised maintenance cost

Capitalised risk S Hnd

Full repair of partial

damage

Only repair of serious

damage(>8%)

No repair of partial

damage

(m) (€) (€) (€)

4 30000 17100 15900

5 3750 1350 1200

5.5 1500 300 300

6 540 90 90

lifetime of 100 years; rate of interest 3.33%

February 1, 2012 19

total cost

Total cost I + S Hnd

Full repair of partial

damage

Only repair of serious

damage(>8%)

No repair of partial

damage

(m) (€) (€) (€)

4 43900 31000 29800

5 18970 16570 16420

5.5 17400 16200 16200

6 17080 16630 16630

6.5 17300

Adding up initial cost plus capitalised maintenance cost

February 1, 2012 20

total cost for various strategies

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

3.5 4 4.5 5 5.5 6 6.5

H-design

co

st

full repair

partly repair

no repair

initial cost

February 1, 2012 21

Conclusions for Rubble Mound Breakwaters

• There is an optimum design wave height

• Accepting regular maintenance is the best option

• This implies that the design also should allow a “repairable” breakwater

February 1, 2012 22

differences in breakwater type

• In case of overload a Rubble mound breakwater will suffer from some damage, which can be repaired.

• In general for Rubble mounds:

the amount of repair costs increase linear with the amount of overload: cost = B * (Hstorm - Hdesign)

• In general for Vertical wall breakwaters:

you have always a given fixed amount of damage, not depending on the amount of overload: cost = A + B (Hstorm - Hdesign)

February 1, 2012 23

Vertical wall breakwater

0

50000

100000

150000

200000

250000

300000

350000

400000

5 7 9 11

Hs (design wave)

tota

l co

st

per

mete

r

building cost

yearly repair

capitalised repair

total

damage cost = 4000 + 1500*Hs

February 1, 2012 24

Conclusions

0

10000

20000

30000

40000

50000

5 7 9 11

Hs (design wave)

tota

l co

st

per

mete

r

0

10000

20000

30000

40000

50000

5 7 9 11

Hs (design wave)

tota

l co

st

per

mete

r

Rubble mound breakwater Vertical wall breakwater

total

repair

February 1, 2012 25

Conclusions (2)

• Rubble mound breakwaters are less sensitive to uncertainties in wave data

• If there is no overload, a Vertical wall breakwater requires less maintenance

• If there is overload, Vertical wall breakwaters cause much more problems

February 1, 2012 26

Including “secondary damage”

• When a breakwater is damaged, the port cannot function well

• The cost because of loss of production should be included in the calculation

• In general secondary damage will not change the tendency of the conclusion before, but make them more even more stronger: • The optimum for a rubble

mound breakwater is allowing quite some damage, and doing a lot of repair